This application requests five years of funding for the Clinical Research Center (MHCRC) for the Study of Suicidal Behavior. These are approximately 30,000 suicides per year in the United States and ten times that number of suicide attempts. To respond to this major health problem, we must improve on current predictors of suicide risk which have high sensitivity but inadequate specificity. Recently identified reductions in serotonin function in the brain of suicide victims and in suicide attempters create hope that more specific predictors of suicide risk may be found. A multidisciplinary approach is necessary to develop a predictive model for suicidal behavior because potential risk factors are demographic, social, developmental, psychiatric, genetic and biological. Cross-sectional identification of risk factors and ultimately prospective testing requires high risk populations where athe frequency of suicidal behavior is sufficient to test risk testing requires high risk populations where the frequency of suicidal behavior is sufficient to test risk factors. Studies of treatment interventions to reduce suicide risk also require specific predictors in order to identify a high risk group. This type of high risk population is generally excluded from biological and treatment studies. This MHCRC has utilized an integrated approach to the multidimensional problem of suicidal behavior. The MHCRC has successfully studies correlates of suicide attempts in 334 patients with major depression, schizophrenia, or borderline personality disorder. In parallel, the MHCRC has studies clinical and biological correlates of suicide in suicide victims and nonsuicide controls. There has been a significant growth i related R01-supported research, development and testing of research instrumentation, training of investigators, and research publicaitons. Four methodological advances distinguish the MHCRC studies of suicidal behavior. First, an integrated series of multi-disciplinary core measures (social, demographic, psychological, psychiatric and biological) evaluate hypothesized risk factors for suicidal behavior and allow estimation of the relative separate and common variance explained by these indices. Second, measures made in different adult patients populations, including affective disorders, personality disorders and schizophrenia and across the life cycle (prepubertal, adolescents and the elderly) will establish which risk factors are disease-specific and age-dependent. Third, a prospective follow-up study is underway in major depression, borderline personality disorder and schizophrenic patients, and in adolescents, to test promising predictors suggested by cross-sectional study. Fourth, results of these clinical studies together with results of the extensive program of concurrent psychological and biological study of completed suicides is being employed to develop an integrated model of suicidal behavior.